| 1 | #region License Information
|
|---|
| 2 | /* HeuristicLab
|
|---|
| 3 | * Copyright (C) 2002-2010 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
|---|
| 4 | *
|
|---|
| 5 | * This file is part of HeuristicLab.
|
|---|
| 6 | *
|
|---|
| 7 | * HeuristicLab is free software: you can redistribute it and/or modify
|
|---|
| 8 | * it under the terms of the GNU General Public License as published by
|
|---|
| 9 | * the Free Software Foundation, either version 3 of the License, or
|
|---|
| 10 | * (at your option) any later version.
|
|---|
| 11 | *
|
|---|
| 12 | * HeuristicLab is distributed in the hope that it will be useful,
|
|---|
| 13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|---|
| 14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|---|
| 15 | * GNU General Public License for more details.
|
|---|
| 16 | *
|
|---|
| 17 | * You should have received a copy of the GNU General Public License
|
|---|
| 18 | * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.
|
|---|
| 19 | */
|
|---|
| 20 | #endregion
|
|---|
| 21 |
|
|---|
| 22 | using System;
|
|---|
| 23 | using HeuristicLab.Analysis;
|
|---|
| 24 | using HeuristicLab.Common;
|
|---|
| 25 | using HeuristicLab.Core;
|
|---|
| 26 | using HeuristicLab.Data;
|
|---|
| 27 | using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
|
|---|
| 28 | using HeuristicLab.Operators;
|
|---|
| 29 | using HeuristicLab.Optimization;
|
|---|
| 30 | using HeuristicLab.Optimization.Operators;
|
|---|
| 31 | using HeuristicLab.Parameters;
|
|---|
| 32 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
|---|
| 33 | using HeuristicLab.Problems.DataAnalysis.Symbolic;
|
|---|
| 34 | using System.Collections.Generic;
|
|---|
| 35 | using HeuristicLab.Problems.DataAnalysis.Symbolic.Symbols;
|
|---|
| 36 | using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding.Symbols;
|
|---|
| 37 |
|
|---|
| 38 | namespace HeuristicLab.Problems.DataAnalysis.Regression.Symbolic.Analyzers {
|
|---|
| 39 | /// <summary>
|
|---|
| 40 | /// An operator that analyzes the population diversity using fine grained structural tree similarity estimation.
|
|---|
| 41 | /// </summary>
|
|---|
| 42 | [Item("FineGrainedStructuralPopulationDiversityAnalyzer", "An operator that analyzes the population diversity using fine grained structural tree similarity estimation.")]
|
|---|
| 43 | [StorableClass]
|
|---|
| 44 | public sealed class FineGrainedStructuralPopulationDiversityAnalyzer : SymbolicRegressionPopulationDiversityAnalyzer {
|
|---|
| 45 |
|
|---|
| 46 | [StorableConstructor]
|
|---|
| 47 | private FineGrainedStructuralPopulationDiversityAnalyzer(bool deserializing) : base(deserializing) { }
|
|---|
| 48 | private FineGrainedStructuralPopulationDiversityAnalyzer(FineGrainedStructuralPopulationDiversityAnalyzer original, Cloner cloner) : base(original, cloner) { }
|
|---|
| 49 | public FineGrainedStructuralPopulationDiversityAnalyzer() : base() { }
|
|---|
| 50 |
|
|---|
| 51 | public override IDeepCloneable Clone(Cloner cloner) {
|
|---|
| 52 | return new FineGrainedStructuralPopulationDiversityAnalyzer(this, cloner);
|
|---|
| 53 | }
|
|---|
| 54 |
|
|---|
| 55 | protected override double[,] CalculateSimilarities(SymbolicExpressionTree[] solutions) {
|
|---|
| 56 | int n = solutions.Length;
|
|---|
| 57 | List<string> variableNames = new List<string>();
|
|---|
| 58 | foreach (StringValue variableName in ProblemData.InputVariables) {
|
|---|
| 59 | variableNames.Add(variableName.Value);
|
|---|
| 60 | }
|
|---|
| 61 | variableNames.Add(ProblemData.TargetVariable.Value);
|
|---|
| 62 | IList<GeneticInformationItem>[] geneticInformationItemsList = new List<GeneticInformationItem>[n];
|
|---|
| 63 | for (int i = 0; i < n; i++) {
|
|---|
| 64 | geneticInformationItemsList[i] = GeneticInformationItem.getGeneticInformationItems(solutions[i].Root, variableNames);
|
|---|
| 65 | }
|
|---|
| 66 | double[,] result = new double[n, n];
|
|---|
| 67 | for (int i = 0; i < n; i++) {
|
|---|
| 68 | for (int j = 0; j < n; j++) {
|
|---|
| 69 | if (i == j)
|
|---|
| 70 | result[i, j] = 1;
|
|---|
| 71 | else
|
|---|
| 72 | result[i, j] = 0;
|
|---|
| 73 | }
|
|---|
| 74 | }
|
|---|
| 75 | return result;
|
|---|
| 76 | }
|
|---|
| 77 |
|
|---|
| 78 |
|
|---|
| 79 | #region private class GeneticInformationItem
|
|---|
| 80 |
|
|---|
| 81 | private class GeneticInformationItem {
|
|---|
| 82 |
|
|---|
| 83 | private Type myAncestorDefinition;
|
|---|
| 84 | public Type AncestorDefinition {
|
|---|
| 85 | get { return myAncestorDefinition; }
|
|---|
| 86 | }
|
|---|
| 87 |
|
|---|
| 88 | private int myAncestorIndex;
|
|---|
| 89 | public int AncestorIndex {
|
|---|
| 90 | get { return myAncestorIndex; }
|
|---|
| 91 | }
|
|---|
| 92 |
|
|---|
| 93 | private Type myDescendantDefinition;
|
|---|
| 94 | public Type DescendantDefinition {
|
|---|
| 95 | get { return myDescendantDefinition; }
|
|---|
| 96 | }
|
|---|
| 97 |
|
|---|
| 98 | private int myLevelDelta;
|
|---|
| 99 | public int LevelDelta {
|
|---|
| 100 | get { return myLevelDelta; }
|
|---|
| 101 | }
|
|---|
| 102 |
|
|---|
| 103 | // not used in HL 3.3
|
|---|
| 104 | // private double myAncestorCoefficient;
|
|---|
| 105 | // public double AncestorCoefficient {
|
|---|
| 106 | // get { return myAncestorCoefficient; }
|
|---|
| 107 | // }
|
|---|
| 108 |
|
|---|
| 109 | // not used in HL 3.3
|
|---|
| 110 | // private double myAncestorVariableIndex;
|
|---|
| 111 | // public double AncestorVariableIndex {
|
|---|
| 112 | // get { return myAncestorVariableIndex; }
|
|---|
| 113 | // }
|
|---|
| 114 |
|
|---|
| 115 | // not used in HL 3.3
|
|---|
| 116 | // private int myAncestorTimeOffset;
|
|---|
| 117 | // public int AncestorTimeOffset {
|
|---|
| 118 | // get { return myAncestorTimeOffset; }
|
|---|
| 119 | // }
|
|---|
| 120 |
|
|---|
| 121 | // not used in HL 3.3
|
|---|
| 122 | // private int myAncestorVariant;
|
|---|
| 123 | // public int AncestorVariant {
|
|---|
| 124 | // get { return myAncestorVariant; }
|
|---|
| 125 | // }
|
|---|
| 126 |
|
|---|
| 127 | private double myDescendantCoefficient;
|
|---|
| 128 | public double DescendantCoefficient {
|
|---|
| 129 | get { return myDescendantCoefficient; }
|
|---|
| 130 | }
|
|---|
| 131 |
|
|---|
| 132 | private double myDescendantVariableIndex;
|
|---|
| 133 | public double DescendantVariableIndex {
|
|---|
| 134 | get { return myDescendantVariableIndex; }
|
|---|
| 135 | }
|
|---|
| 136 |
|
|---|
| 137 | private int myDescendantTimeOffset;
|
|---|
| 138 | public int DescendantTimeOffset {
|
|---|
| 139 | get { return myDescendantTimeOffset; }
|
|---|
| 140 | }
|
|---|
| 141 |
|
|---|
| 142 | // not used in HL 3.3
|
|---|
| 143 | //private int myDescendantVariant;
|
|---|
| 144 | //public int DescendantVariant {
|
|---|
| 145 | // get { return myDescendantVariant; }
|
|---|
| 146 | //}
|
|---|
| 147 |
|
|---|
| 148 | /*
|
|---|
| 149 | public static GeneticInformationItem FindBestPendant(GeneticInformationItem Item, List<GeneticInformationItem> ComparisonItems,
|
|---|
| 150 | StructuralSimilarityAnalysisParameters Parameters,
|
|---|
| 151 | int MaxTreeHeight, int MaxTimeOffset,
|
|---|
| 152 | out double BestPendantSimilarity) {
|
|---|
| 153 | int maxSimilarityIndex = -1;
|
|---|
| 154 | double similarity, maxSimilarity = -double.MaxValue;
|
|---|
| 155 | for (int i = 0; i < ComparisonItems.Count; i++) {
|
|---|
| 156 | similarity = Similarity(Item, ComparisonItems[i], Parameters, MaxTreeHeight, MaxTimeOffset);
|
|---|
| 157 | if (!double.IsNaN(similarity) && similarity > maxSimilarity) {
|
|---|
| 158 | maxSimilarity = similarity;
|
|---|
| 159 | maxSimilarityIndex = i;
|
|---|
| 160 | }
|
|---|
| 161 | }
|
|---|
| 162 | BestPendantSimilarity = maxSimilarity;
|
|---|
| 163 | if (maxSimilarityIndex >= 0)
|
|---|
| 164 | return ComparisonItems[maxSimilarityIndex];
|
|---|
| 165 | else
|
|---|
| 166 | return null;
|
|---|
| 167 | }*/
|
|---|
| 168 |
|
|---|
| 169 | /*
|
|---|
| 170 | public static double Similarity(GeneticInformationItem Item1, GeneticInformationItem Item2,
|
|---|
| 171 | StructuralSimilarityAnalysisParameters Parameters,
|
|---|
| 172 | int MaxTreeHeight, int MaxTimeOffset) {
|
|---|
| 173 |
|
|---|
| 174 | if (Item1.AncestorDefinition != Item2.AncestorDefinition ||
|
|---|
| 175 | Item1.DescendantDefinition != Item2.DescendantDefinition)
|
|---|
| 176 | return double.NaN;
|
|---|
| 177 |
|
|---|
| 178 | // the two items for sure have the same behavior definitions
|
|---|
| 179 | #region init
|
|---|
| 180 | double punishmentContributionSum = 0;
|
|---|
| 181 | double punishmentCoefficientsProduct = 1;
|
|---|
| 182 | double ancestorCoefficientDifferencePunishment = 0;
|
|---|
| 183 | double ancestorTimeOffsetDifferencePunishment = 0;
|
|---|
| 184 | double ancestorVariantDifferencePunishment = 0;
|
|---|
| 185 | double ancestorVariableIndexDifferencePunishment = 0;
|
|---|
| 186 | double descendantCoefficientDifferencePunishment = 0;
|
|---|
| 187 | double descendantTimeOffsetDifferencePunishment = 0;
|
|---|
| 188 | double descendantVariantDifferencePunishment = 0;
|
|---|
| 189 | double descendantVariableIndexDifferencePunishment = 0;
|
|---|
| 190 | double ancestorIndexDifferencePunishment = 0;
|
|---|
| 191 | double levelDifferencePunishment = 0;
|
|---|
| 192 | #endregion
|
|---|
| 193 |
|
|---|
| 194 | ITerminalDefinition ancestorTerminal = Item1.AncestorDefinition as ITerminalDefinition;
|
|---|
| 195 | ITerminalDefinition descendantTerminal = Item1.DescendantDefinition as ITerminalDefinition;
|
|---|
| 196 | IFunctionDefinition ancestorFunction = Item1.AncestorDefinition as IFunctionDefinition;
|
|---|
| 197 | IFunctionDefinition descendantFunction = Item1.DescendantDefinition as IFunctionDefinition;
|
|---|
| 198 |
|
|---|
| 199 | if (Parameters.ConsiderLevelDifference) {
|
|---|
| 200 | levelDifferencePunishment = Item1.LevelDelta - Item2.LevelDelta;
|
|---|
| 201 | if (levelDifferencePunishment < 0)
|
|---|
| 202 | levelDifferencePunishment = -levelDifferencePunishment;
|
|---|
| 203 | levelDifferencePunishment /= MaxTreeHeight;
|
|---|
| 204 | if (levelDifferencePunishment > 1)
|
|---|
| 205 | levelDifferencePunishment = 1;
|
|---|
| 206 | levelDifferencePunishment *= Parameters.LevelDifferenceCoefficient;
|
|---|
| 207 | punishmentContributionSum += Parameters.LevelDifferenceCoefficient;
|
|---|
| 208 | punishmentCoefficientsProduct *= (1 - Parameters.LevelDifferenceCoefficient);
|
|---|
| 209 | }
|
|---|
| 210 | if (Parameters.ConsiderAncestorIndex) {
|
|---|
| 211 | if (Item1.AncestorIndex != Item2.AncestorIndex)
|
|---|
| 212 | ancestorIndexDifferencePunishment = 1;
|
|---|
| 213 | else
|
|---|
| 214 | ancestorIndexDifferencePunishment = 0;
|
|---|
| 215 | ancestorIndexDifferencePunishment *= Parameters.AncestorIndexCoefficient;
|
|---|
| 216 | punishmentContributionSum += Parameters.AncestorIndexCoefficient;
|
|---|
| 217 | punishmentCoefficientsProduct *= (1 - Parameters.AncestorIndexCoefficient);
|
|---|
| 218 | }
|
|---|
| 219 |
|
|---|
| 220 | if (Item1.AncestorDefinition is ITerminalDefinition) {
|
|---|
| 221 | if (Parameters.ConsiderCoefficient) {
|
|---|
| 222 | double coefficientDifference = Math.Abs(Item1.myAncestorCoefficient - Item2.myAncestorCoefficient);
|
|---|
| 223 | if (ancestorTerminal.Parametrization.CoefficientIsGaussian)
|
|---|
| 224 | ancestorCoefficientDifferencePunishment = (coefficientDifference / (ancestorTerminal.Parametrization.CoefficientSigma * 4));
|
|---|
| 225 | else
|
|---|
| 226 | ancestorCoefficientDifferencePunishment = (coefficientDifference / (ancestorTerminal.Parametrization.CoefficientMaxValue - ancestorTerminal.Parametrization.CoefficientMinValue));
|
|---|
| 227 | if (ancestorCoefficientDifferencePunishment > 1)
|
|---|
| 228 | ancestorCoefficientDifferencePunishment = 1;
|
|---|
| 229 | ancestorCoefficientDifferencePunishment *= Parameters.CoefficientCoefficient;
|
|---|
| 230 | punishmentContributionSum += Parameters.CoefficientCoefficient;
|
|---|
| 231 | punishmentCoefficientsProduct *= (1 - Parameters.CoefficientCoefficient);
|
|---|
| 232 | }
|
|---|
| 233 | if (Parameters.ConsiderTimeOffset) {
|
|---|
| 234 | double timeOffsetDifference = Math.Abs(Item1.AncestorTimeOffset - Item2.AncestorTimeOffset);
|
|---|
| 235 | if (MaxTimeOffset > 0)
|
|---|
| 236 | ancestorTimeOffsetDifferencePunishment = timeOffsetDifference / MaxTimeOffset;
|
|---|
| 237 | ancestorTimeOffsetDifferencePunishment *= Parameters.TimeOffsetCoefficient;
|
|---|
| 238 | punishmentContributionSum += Parameters.TimeOffsetCoefficient;
|
|---|
| 239 | punishmentCoefficientsProduct *= (1 - Parameters.TimeOffsetCoefficient);
|
|---|
| 240 | }
|
|---|
| 241 | if (Parameters.ConsiderVariableIndex) {
|
|---|
| 242 | if (Item1.AncestorVariableIndex != Item2.AncestorVariableIndex)
|
|---|
| 243 | ancestorVariableIndexDifferencePunishment = 1;
|
|---|
| 244 | else
|
|---|
| 245 | ancestorVariableIndexDifferencePunishment = 0;
|
|---|
| 246 | ancestorVariableIndexDifferencePunishment *= Parameters.VariableIndexCoefficient;
|
|---|
| 247 | punishmentContributionSum += Parameters.VariableIndexCoefficient;
|
|---|
| 248 | punishmentCoefficientsProduct *= (1 - Parameters.VariableIndexCoefficient);
|
|---|
| 249 | }
|
|---|
| 250 | } else {
|
|---|
| 251 | if (Parameters.ConsiderVariant) {
|
|---|
| 252 | if (Item1.AncestorVariant != Item2.AncestorVariant)
|
|---|
| 253 | ancestorVariantDifferencePunishment = 1;
|
|---|
| 254 | else
|
|---|
| 255 | ancestorVariantDifferencePunishment = 0;
|
|---|
| 256 | ancestorVariantDifferencePunishment *= Parameters.VariantCoefficient;
|
|---|
| 257 | punishmentContributionSum += Parameters.VariantCoefficient;
|
|---|
| 258 | punishmentCoefficientsProduct *= (1 - Parameters.VariantCoefficient);
|
|---|
| 259 | }
|
|---|
| 260 | }
|
|---|
| 261 |
|
|---|
| 262 | if (Item1.DescendantDefinition is ITerminalDefinition) {
|
|---|
| 263 | if (Parameters.ConsiderCoefficient) {
|
|---|
| 264 | double coefficientDifference = Math.Abs(Item1.myDescendantCoefficient - Item2.myDescendantCoefficient);
|
|---|
| 265 | if (descendantTerminal.Parametrization.CoefficientIsGaussian)
|
|---|
| 266 | descendantCoefficientDifferencePunishment = (coefficientDifference / (descendantTerminal.Parametrization.CoefficientSigma * 4));
|
|---|
| 267 | else
|
|---|
| 268 | descendantCoefficientDifferencePunishment = (coefficientDifference / (descendantTerminal.Parametrization.CoefficientMaxValue - descendantTerminal.Parametrization.CoefficientMinValue));
|
|---|
| 269 | if (descendantCoefficientDifferencePunishment > 1)
|
|---|
| 270 | descendantCoefficientDifferencePunishment = 1;
|
|---|
| 271 | descendantCoefficientDifferencePunishment *= Parameters.CoefficientCoefficient;
|
|---|
| 272 | punishmentContributionSum += Parameters.CoefficientCoefficient;
|
|---|
| 273 | punishmentCoefficientsProduct *= (1 - Parameters.CoefficientCoefficient);
|
|---|
| 274 | }
|
|---|
| 275 | if (Parameters.ConsiderTimeOffset) {
|
|---|
| 276 | double timeOffsetDifference = Math.Abs(Item1.DescendantTimeOffset - Item2.DescendantTimeOffset);
|
|---|
| 277 | if (MaxTimeOffset > 0)
|
|---|
| 278 | descendantTimeOffsetDifferencePunishment = timeOffsetDifference / MaxTimeOffset;
|
|---|
| 279 | descendantTimeOffsetDifferencePunishment *= Parameters.TimeOffsetCoefficient;
|
|---|
| 280 | punishmentContributionSum += Parameters.TimeOffsetCoefficient;
|
|---|
| 281 | punishmentCoefficientsProduct *= (1 - Parameters.TimeOffsetCoefficient);
|
|---|
| 282 | }
|
|---|
| 283 | if (Parameters.ConsiderVariableIndex) {
|
|---|
| 284 | if (Item1.DescendantVariableIndex != Item2.DescendantVariableIndex)
|
|---|
| 285 | descendantVariableIndexDifferencePunishment = 1;
|
|---|
| 286 | else
|
|---|
| 287 | descendantVariableIndexDifferencePunishment = 0;
|
|---|
| 288 | descendantVariableIndexDifferencePunishment *= Parameters.VariableIndexCoefficient;
|
|---|
| 289 | punishmentContributionSum += Parameters.VariableIndexCoefficient;
|
|---|
| 290 | punishmentCoefficientsProduct *= (1 - Parameters.VariableIndexCoefficient);
|
|---|
| 291 | }
|
|---|
| 292 | } else {
|
|---|
| 293 | if (Parameters.ConsiderVariant) {
|
|---|
| 294 | if (Item1.DescendantVariant != Item2.DescendantVariant)
|
|---|
| 295 | descendantVariantDifferencePunishment = 1;
|
|---|
| 296 | else
|
|---|
| 297 | descendantVariantDifferencePunishment = 0;
|
|---|
| 298 | descendantVariantDifferencePunishment *= Parameters.VariantCoefficient;
|
|---|
| 299 | punishmentContributionSum += Parameters.VariantCoefficient;
|
|---|
| 300 | punishmentCoefficientsProduct *= (1 - Parameters.VariantCoefficient);
|
|---|
| 301 | }
|
|---|
| 302 | }
|
|---|
| 303 |
|
|---|
| 304 | double result;
|
|---|
| 305 |
|
|---|
| 306 | if (Parameters.AdditiveSimilarityCalculation) {
|
|---|
| 307 | double punishmentsSum =
|
|---|
| 308 | ancestorCoefficientDifferencePunishment + ancestorTimeOffsetDifferencePunishment +
|
|---|
| 309 | ancestorVariantDifferencePunishment + ancestorVariableIndexDifferencePunishment +
|
|---|
| 310 | descendantCoefficientDifferencePunishment + descendantTimeOffsetDifferencePunishment +
|
|---|
| 311 | descendantVariantDifferencePunishment + descendantVariableIndexDifferencePunishment +
|
|---|
| 312 | ancestorIndexDifferencePunishment + levelDifferencePunishment;
|
|---|
| 313 | result = (1 - punishmentsSum / punishmentContributionSum);
|
|---|
| 314 | } else {
|
|---|
| 315 | result =
|
|---|
| 316 | (1 - ancestorCoefficientDifferencePunishment) *
|
|---|
| 317 | (1 - ancestorTimeOffsetDifferencePunishment) *
|
|---|
| 318 | (1 - ancestorVariantDifferencePunishment) *
|
|---|
| 319 | (1 - ancestorVariableIndexDifferencePunishment) *
|
|---|
| 320 | (1 - descendantCoefficientDifferencePunishment) *
|
|---|
| 321 | (1 - descendantTimeOffsetDifferencePunishment) *
|
|---|
| 322 | (1 - descendantVariantDifferencePunishment) *
|
|---|
| 323 | (1 - descendantVariableIndexDifferencePunishment) *
|
|---|
| 324 | (1 - ancestorIndexDifferencePunishment) *
|
|---|
| 325 | (1 - levelDifferencePunishment);
|
|---|
| 326 | // worst possible result is (1-punishmentCoefficientsProduct), so scale linearly to [0;1]:
|
|---|
| 327 | result = (result - punishmentCoefficientsProduct) / (1 - punishmentCoefficientsProduct);
|
|---|
| 328 | }
|
|---|
| 329 |
|
|---|
| 330 | if (result < 0 || result > 1)
|
|---|
| 331 | throw new InvalidOperationException("ERROR in GeneticInformationItem.Similarity: An invalid similarity value (" + result.ToString() + ") has been calculated.");
|
|---|
| 332 |
|
|---|
| 333 | return result;
|
|---|
| 334 |
|
|---|
| 335 | }*/
|
|---|
| 336 |
|
|---|
| 337 | /*
|
|---|
| 338 | public static List<GeneticInformationItem> GetGeneticInformationItems(SymbolicExpressionTreeNode Formula, int MinLevelDifference, int MaxLevelDifference) {
|
|---|
| 339 | List<GeneticInformationItem> result = new List<GeneticInformationItem>();
|
|---|
| 340 | List<SymbolicExpressionTreeNode> allNodes = new List<SymbolicExpressionTreeNode>();
|
|---|
| 341 | allNodes.Add(Formula);
|
|---|
| 342 | allNodes.AddRange(getDescendants(Formula));
|
|---|
| 343 | foreach (SymbolicExpressionTreeNode node in allNodes) {
|
|---|
| 344 | List<SymbolicExpressionTreeNode> allDescendants = new List<SymbolicExpressionTreeNode>();
|
|---|
| 345 | allDescendants.Add(node);
|
|---|
| 346 | allDescendants.AddRange(getDescendants(node));
|
|---|
| 347 | foreach (SymbolicExpressionTreeNode descendant in allDescendants) {
|
|---|
| 348 | GeneticInformationItem item = new GeneticInformationItem(node, descendant);
|
|---|
| 349 | if (item.LevelDelta >= MinLevelDifference && item.LevelDelta <= MaxLevelDifference)
|
|---|
| 350 | result.Add(item);
|
|---|
| 351 | }
|
|---|
| 352 | }
|
|---|
| 353 | return result;
|
|---|
| 354 | }*/
|
|---|
| 355 |
|
|---|
| 356 | /*
|
|---|
| 357 | public static List<SymbolicExpressionTreeNode> GetDescendants(SymbolicExpressionTreeNode node) {
|
|---|
| 358 | List<SymbolicExpressionTreeNode> descendants = new List<SymbolicExpressionTreeNode>();
|
|---|
| 359 | foreach (SymbolicExpressionTreeNode subTreeNode in node.SubTrees) {
|
|---|
| 360 | AddDescendants(descendants, subTreeNode);
|
|---|
| 361 | }
|
|---|
| 362 | return descendants;
|
|---|
| 363 | }
|
|---|
| 364 | private static void AddDescendants(List<SymbolicExpressionTreeNode> list, SymbolicExpressionTreeNode node) {
|
|---|
| 365 | list.Add(node);
|
|---|
| 366 | foreach (SymbolicExpressionTreeNode subTreeNode in node.SubTrees) {
|
|---|
| 367 | AddDescendants(list, subTreeNode);
|
|---|
| 368 | }
|
|---|
| 369 | }*/
|
|---|
| 370 |
|
|---|
| 371 | public static IList<GeneticInformationItem> getGeneticInformationItems(SymbolicExpressionTreeNode node, List<string> variableNames) {
|
|---|
| 372 | // Idea: collect all descendants' lists and then add new items using the retrieved ones.
|
|---|
| 373 | // This should save lots of time and reduce complexity of the items retrieval process.
|
|---|
| 374 | if (node.Symbol is ProgramRootSymbol)
|
|---|
| 375 | return getGeneticInformationItems(node.SubTrees[0], variableNames);
|
|---|
| 376 | List<GeneticInformationItem> list = new List<GeneticInformationItem>();
|
|---|
| 377 | // add item for this node:
|
|---|
| 378 | list.Add(new GeneticInformationItem(node, variableNames));
|
|---|
| 379 | // add items for the descendants:
|
|---|
| 380 | for (int i = 0; i < node.SubTrees.Count; i++) {
|
|---|
| 381 | IList<GeneticInformationItem> descendantItems = getGeneticInformationItems(node.SubTrees[i], variableNames);
|
|---|
| 382 | list.AddRange(descendantItems);
|
|---|
| 383 | foreach (GeneticInformationItem item in descendantItems) {
|
|---|
| 384 | list.Add(new GeneticInformationItem(node, item, i));
|
|---|
| 385 | }
|
|---|
| 386 | }
|
|---|
| 387 | return list;
|
|---|
| 388 | }
|
|---|
| 389 |
|
|---|
| 390 | private GeneticInformationItem (SymbolicExpressionTreeNode node, List<string> variableNames) {
|
|---|
| 391 | myAncestorIndex = -1;
|
|---|
| 392 | VariableTreeNode variableTreeNode = node as VariableTreeNode;
|
|---|
| 393 | LaggedVariableTreeNode laggedVariableTreeNode = node as LaggedVariableTreeNode;
|
|---|
| 394 | ConstantTreeNode constantTreeNode = node as ConstantTreeNode;
|
|---|
| 395 | myAncestorDefinition = node.Symbol.GetType();
|
|---|
| 396 | myDescendantDefinition = myAncestorDefinition;
|
|---|
| 397 | if (variableTreeNode != null)
|
|---|
| 398 | myDescendantCoefficient = variableTreeNode.Weight;
|
|---|
| 399 | else if (constantTreeNode != null)
|
|---|
| 400 | myDescendantCoefficient = constantTreeNode.Value;
|
|---|
| 401 | else
|
|---|
| 402 | myDescendantCoefficient = double.NaN;
|
|---|
| 403 | if (laggedVariableTreeNode != null)
|
|---|
| 404 | myDescendantTimeOffset = laggedVariableTreeNode.Lag;
|
|---|
| 405 | else
|
|---|
| 406 | myDescendantTimeOffset = 0;
|
|---|
| 407 | if (variableTreeNode != null)
|
|---|
| 408 | myDescendantVariableIndex = variableNames.IndexOf(variableTreeNode.VariableName);
|
|---|
| 409 | else
|
|---|
| 410 | myDescendantVariableIndex = -1;
|
|---|
| 411 | myLevelDelta = 0;
|
|---|
| 412 | }
|
|---|
| 413 |
|
|---|
| 414 | private GeneticInformationItem(SymbolicExpressionTreeNode parentNode, GeneticInformationItem descendantGeneticInformationItem, int ancestorIndex) {
|
|---|
| 415 | myAncestorIndex = ancestorIndex;
|
|---|
| 416 | myLevelDelta = descendantGeneticInformationItem.LevelDelta + 1;
|
|---|
| 417 | myAncestorDefinition = parentNode.Symbol.GetType();
|
|---|
| 418 | myDescendantCoefficient = descendantGeneticInformationItem.DescendantCoefficient;
|
|---|
| 419 | myDescendantDefinition = descendantGeneticInformationItem.DescendantDefinition;
|
|---|
| 420 | myDescendantTimeOffset = descendantGeneticInformationItem.DescendantTimeOffset;
|
|---|
| 421 | myDescendantVariableIndex = descendantGeneticInformationItem.DescendantVariableIndex;
|
|---|
| 422 | }
|
|---|
| 423 |
|
|---|
| 424 | public override string ToString() {
|
|---|
| 425 | return "ancestor: " + AncestorDefinition.Name.ToString() + ", [" + AncestorIndex + "], delta " + LevelDelta
|
|---|
| 426 | + "; descendant: " + DescendantDefinition.Name.ToString() + " (varIndex " + DescendantVariableIndex + ", "
|
|---|
| 427 | + DescendantCoefficient + ", t-" + DescendantTimeOffset + ")";
|
|---|
| 428 | }
|
|---|
| 429 |
|
|---|
| 430 | }
|
|---|
| 431 |
|
|---|
| 432 | #endregion
|
|---|
| 433 |
|
|---|
| 434 | }
|
|---|
| 435 | }
|
|---|
